Currently, a networking mode of device nodes in a common access network basically adopts a distributed autonomic network structure. In this network structure, each network node is configured separately and works independently. Therefore, asymmetric information between network nodes exists, which causes problems such as incapability of sharing service capabilities between different network nodes, lack of collaboration, and redundant and repeated execution of service functions, so that the processing performance of the whole network is poor. Several scenarios of unreasonable network node capability distribution that are very common in network application are described as follows:
1) As shown in FIG. 1, multiple client clusters and server clusters access a network through different network nodes, where a client cluster 1 accesses the network through a node A, and a server cluster 1 accesses the network through a node B. An HTTP (Hypertext Transport Protocol) protocol-based IPS (Intrusion Prevention System) service processing capability is allocated for the network node A; and a service processing capability of URL (Uniform Resource Locator) filtering is allocated for the node B. When a network service flow passes the node A and the node B sequentially, the node A and the node B perform DPI (Deep Packet Inspection) identification and parsing processing on the service flow at an application layer due to the need of service processing, thereby causing that different devices execute a part of functions repeatedly on the same service flow.
2) As shown in FIG. 2, multiple client clusters and server clusters access a network through different network nodes, where a client cluster 1 accesses the network through a node A, and a server cluster 1 accesses the network through a node B. A message content-based compression capability is allocated for the network node A, but the node B does not have the decompression capability. When transmission acceleration needs to be performed on a service flow at an application layer the message may undergo content compression when passing the source end node A, but cannot be decompressed when passing the destination node B, so that a network acceleration service cannot be implemented.
Due to the distributed structure of the existing network and the separate deployment manner of nodes, service capabilities of network nodes are private and in lack of unified collaboration management, so that the whole network lacks collaboration for related service processing at the application layer, the processing is redundant, and the efficiency is reduced.
How to implement unified resource management on network nodes, reasonably distribute node capabilities and coordinate service scheduling, and implement multi-node capability sharing and cooperation so as to improve the processing efficiency of the whole network is a major problem being faced currently.